Analysis of social interactions and prediction of assignment grades in a Massive Open Online Course
MOOCs (Massive Open Online Courses) have been gaining interest in the last years and they provide a lot of data for analysis which can be exploited applying learning analytics techniques. In these courses, social interactions mainly occur in the course forum and consequently, there is an interest in mining forum data to discover learners behaviors. Because of that, this work presents a dashboard designed to provide visualizations related to forum interactions on edX courses and discusses the results obtained in a real MOOC, including sentiment analysis. Furthermore, as it is also important to identify learners performance to improve the learning processes, there is also an analysis of how assignment grades can be predicted and how much they can be anticipated. Besides, the effect of different indicators on grades and particularly the relationship with forum interactions is discussed.
Pedro Manuel Moreno Marcos Universidad Carlos III, Madrid